JPWO2015052842A1 - Mass spectrometry data analyzer - Google Patents

Abstract

After reading the mass spectrum data obtained at all measurement points in the two-dimensional measurement region (S1), for each m / z value, the spatial density level dependent method for the image based on the data collected from all measurement points Statistical texture analysis such as the above is executed to calculate the feature amount (S2 to S5). Then, a spectrum is created from the feature quantity for each m / z value (S6), peak detection is performed on the spectrum, a predetermined number of peaks are selected as analysis target peaks in order of intensity, and the m / z value is obtained (S7). , S8). Statistical texture analysis finds features that reflect the specific spatial distribution of materials, so it is possible to select m / z values that are useful for the analysis of specific spatial distributions, and data with such m / z values. By selecting and using for multivariate analysis, it is possible to extract more useful information than conventional from mass spectrometry imaging data.

Description

  The present invention analyzes mass analysis imaging data collected by performing mass analysis on each of a plurality of measurement points in a measurement region on a sample, and analyzes the two-dimensional distribution of substances contained in the sample. The present invention relates to a mass spectrometry data analysis apparatus for acquiring significant information.

  A device called a mass microscope or an imaging mass spectrometer has been developed as a device for observing the morphology of a sample such as a biological tissue and simultaneously measuring the distribution of a substance (molecule) present in a predetermined region on the sample. . According to such an apparatus, a specific measurement region included in an arbitrary measurement region on a specified sample based on an image obtained by microscopic observation is maintained while maintaining the shape of the sample almost without crushing or crushing the sample. It is possible to obtain a two-dimensional distribution image (mass spectrometry imaging image) of ions having a mass-to-charge ratio m / z. Therefore, in particular, in the biochemical field, the medical / pharmaceutical field, and the like, applications such as obtaining distribution information of components such as specific proteins contained in in vivo cells are expected.

In the analysis by the imaging mass spectrometer, mass spectrum data indicating the relationship between the mass-to-charge ratio and the signal intensity over a predetermined mass-to-charge ratio range is obtained for each measurement point in the two-dimensional measurement region. In addition, when performing MS ⁿ analysis in an imaging mass spectrometer, MS ⁿ spectrum data indicating the relationship between the mass-to-charge ratio of product ions with respect to one or a plurality of precursor ions and signal intensity is obtained for each measurement point. For this reason, the amount of data obtained in the entire sample or the entire measurement region on the sample is extremely large. It would be impractical for humans to extract meaningful information visually from such a vast amount of data, requiring a lot of time and effort.

  For example, in measurement of a biological sample using an imaging mass spectrometer, it is often important to specify a substance that is specifically distributed on the sample. Therefore, principal component analysis (PCA = Principal Component Analysis), hierarchical cluster analysis (HCA = Hierarchical Cluster Analysis), etc., are used to determine the mass-to-charge ratio of ions that are specifically distributed within a two-dimensional measurement region. A method using multivariate analysis has been reported (see Non-Patent Document 1, for example).

  The peaks appearing in the mass spectrum obtained by mass spectrometry include peaks derived from a large number of noises, and peaks that are derived from components contained in the sample but are not significant. Therefore, when the multivariate analysis process as described above is executed, an appropriate number of analysis target peaks are obtained by performing a peak extraction process on the collected mass spectrometry imaging data prior to the process. It is common to elect. Then, data corresponding to the mass-to-charge ratio value of the analysis target peak is extracted from the mass spectrum data at each measurement point, and the spectrum intensity at each measurement point is determined based on this data. As a result, noise peaks and the like can be eliminated as much as possible, significant information can be extracted, and processing time can be shortened by reducing data to be processed.

  As a peak extraction method for selecting an appropriate analysis target peak, the method described in Non-Patent Document 2 is well known. In this peak extraction method, analysis target peaks are selected mainly by selecting a specified number of peaks in descending order of signal intensity in two types of spectra. One of these two types of spectra is a spectrum created by integrating the signal intensity values in the mass spectrum at all measurement points in the measurement region for each mass-to-charge ratio value, or dividing this by the number of measurement points. It is the spectrum which calculated | required the average value of a signal strength value. Hereinafter, this spectrum is referred to as an average spectrum. The other one is a mass spectrum obtained by extracting a peak showing the maximum intensity for each mass-to-charge ratio value from the mass spectrum at all measurement points in the measurement region, and collecting and reconstructing the peak. Hereinafter, this mass spectrum is referred to as a base peak spectrum. For example, FIG. 4 of Non-Patent Document 2 discloses an example of an average spectrum (mean spectrum) and a base peak spectrum (basepeak spectrum).

Gregor McCombie and three others, “Spatial and Spectral Correlations in MALDI Mass Spectrometry Images by Clustering and Multivariate Analysis ”, Analytical Chemistry, Vol. 77, 2005, pp. 6118-6124 Liam A. McDonnell and four others, “Imaging Mass Spectrometry Data Reduction: Automated Feature Identification and Extraction), Journal of American Society for Mass Spectrometry, 2010, Vol. 21, pp.1969-1978 “Analyzing the Texture of an Image”, [searched September 30, 2013], USA, MathWorks, Internet <URL: http: // www.mathworks.co.jp/en/help/images/analyzing-the-texture-of-an-image.html?lang=en> Haralick RM, “Textural Features for Image Classification”, iTriplee Transactions on Systems, Mann and Cybernetics (IEEE Transactions on Systems) , Man and Cybernetics), SMC-3, 1973, pp.610-621

  According to the peak extraction method using the above-described average spectrum and base peak spectrum, there is a high possibility that an ion-derived peak showing a relatively large signal intensity appears at a higher intensity among all measurement points. However, when creating an average spectrum or a base peak spectrum, the spatial distribution as mass spectrometry imaging data is not considered. That is, in this conventional peak extraction method, a peak having a large signal intensity is preferentially selected as the entire measurement point. However, a substance corresponding to a peak showing a large signal intensity does not necessarily have a specific spatial distribution. For example, when the same substance is contained in a relatively large amount in the entire measurement region, a peak derived from the substance is selected as an analysis target peak, but is not useful for extracting significant information. Therefore, when the conventional peak extraction method is used, there is a possibility that a substance intended by the analyst, that is, a substance that is specifically distributed, may not be appropriately extracted.

  The present invention has been made to solve the above-mentioned problems, and the object of the present invention is that the signal intensity per measurement point is relatively large in the data collected by imaging mass spectrometry, and Provided is a mass spectrometry data analysis apparatus capable of extracting significant information about a substance that is specifically distributed by using a peak extraction method capable of extracting a peak group having a specific spatial distribution. There is.

In order to solve the above-described problems, the present invention provides mass spectral data collected by performing mass spectrometry on a plurality of measurement points set in a two-dimensional measurement region on a sample. A mass spectrometry data analysis device for analysis processing,
a) Based on the mass spectrum data for each measurement point in the measurement region, texture analysis is performed for each image showing a two-dimensional distribution of signal intensity values created for each mass-charge value within a predetermined mass-to-charge ratio range. A texture analysis execution unit for obtaining a feature amount;
b) a feature amount spectrum creation unit that creates a mass spectrum based on the feature amount obtained for each mass-to-charge ratio value by the texture analysis execution unit;
c) an analysis target peak determination unit that performs peak detection on the feature amount spectrum created by the feature amount spectrum creation unit, and extracts a predetermined number of peaks as an analysis target peak in order of intensity among the obtained plurality of peaks; ,
And selecting and processing data from the mass spectrum data at each measurement point in the measurement region with reference to the analysis target peak determined by the analysis target peak determination unit.

The mass spectrum data to be analyzed in the mass spectrometry data analysis apparatus according to the present invention includes MS ⁿ analysis (where n is an integer of 2 or more) in addition to data obtained by performing normal mass analysis on each measurement point. ) Is also included.

  The mass spectrometry data analysis apparatus according to the present invention uses image texture analysis often used for image matching, region classification, and the like in order to find a mass-to-charge ratio in which signal intensity is spatially and specifically distributed. There are two types of texture analysis: structural texture analysis and statistical texture analysis. In biological samples that are often measured by imaging mass spectrometry, in many cases, the structure and rules of luminance and color patterns in the measurement region Sex is scarce. Since the statistical texture analysis is suitable for the texture analysis for such an image, the texture analysis execution unit preferably executes the statistical texture analysis in the present invention.

In the mass spectrometry data analysis apparatus according to the present invention, when mass spectrum data for each measurement point in the measurement region is given as a processing target, the texture analysis execution unit determines whether or not all measurement points are within a predetermined mass-to-charge ratio range. The data indicating the signal intensity with respect to the mass charge value is extracted, and texture analysis is performed on the image indicating the two-dimensional distribution of the signal intensity value, thereby calculating the texture feature amount of the image. Then, by executing such processing for all the mass-to-charge values within a predetermined mass-to-charge ratio range, the texture feature amounts for all the mass-to-charge ratio values are obtained.
Note that the predetermined mass-to-charge ratio range (the upper limit value and the lower limit value) may be set in advance as default values, for example, or may be designated by the analyst each time.

  The feature amount spectrum creating unit creates a feature amount spectrum indicating the relationship between the mass to charge ratio and the texture feature amount based on the feature amount obtained for each mass to charge ratio value as described above. The analysis target peak determination unit performs peak detection on the feature amount spectrum instead of the above-described average spectrum and base peak spectrum. Then, a predetermined number of peaks are extracted as analysis target peaks in the order of intensity (that is, in descending order of texture feature amount) from among the obtained many peaks. Thereby, a predetermined number of mass-to-charge ratios to be analyzed are obtained. Therefore, for example, data corresponding to the mass-to-charge ratio to be analyzed is selected from the mass spectrum data at each measurement point in the measurement region, and the data thus selected is used for appropriate multivariate analysis processing, for example. By providing, information on a substance that is spatially specifically distributed is calculated. In the mass spectrometry data analysis apparatus according to the present invention, the method for processing data selected from the mass spectrum data at each measurement point is not particularly limited.

  In texture analysis, information on patterns and patterns existing in an image is obtained, which can be said to reflect a spatial distribution. In particular, in statistical texture analysis, various feature quantities (statistics) reflecting local patterns and pattern features in an image can be obtained as analysis results. Therefore, by using an appropriate feature amount, it is possible to extract a mass-to-charge ratio such that the signal intensity is relatively large in the entire measurement region and the ion spatial distribution is specific. There are also several analysis methods for statistical texture analysis, and it is preferable to select an appropriate analysis method and use an appropriate feature amount.

  According to an experimental study by the present inventor, a spatial density level dependent method (SGLDM = Spatial Gray Level Dependant Method) or a density level differential method (GLDM = Gray Level Differential Method) is particularly effective as statistical texture analysis. is there. In the case of the spatial density level dependent method, a correlation value may be used as a feature value, and in the case of the density level difference method, a contrast value may be used as a feature value. Of course, other statistical texture analysis methods and feature quantities may be used.

  According to the mass spectrometry data analysis apparatus according to the present invention, the mass-to-charge ratio of ions derived from substances that are specifically distributed in the measurement region, compared to the conventional peak selection method using the average spectrum or the base peak spectrum. Can be found accurately. Therefore, for example, by performing multivariate analysis using mass spectrometry imaging data corresponding to a plurality of extracted peaks, the mass-to-charge ratio value of a substance localized in a specific site that could not be found in the past is obtained. The mass spectrometric imaging image at the mass to charge ratio value can be identified and provided to the analyst.

BRIEF DESCRIPTION OF THE DRAWINGS The schematic block diagram of one Example of the imaging mass spectrometer using the mass spectrometry data analyzer which concerns on this invention. BRIEF DESCRIPTION OF THE DRAWINGS Schematic explanatory drawing of the characteristic data analysis processing operation in the imaging mass spectrometer of a present Example. The flowchart of the characteristic data analysis process in the imaging mass spectrometer of a present Example. The figure which shows the mass extraction imaging image with respect to the peak extraction result in the measurement example 1, and the mass to charge ratio value of each peak. The figure which shows the spectrum used for the peak extraction in the measurement example 1. FIG. The figure which shows the mass spectrometry imaging image with respect to the mass extraction ratio value of the peak extraction result in each measurement example 2, and each peak. The figure which shows the spectrum used for the peak extraction in the measurement example 2. FIG.

  Hereinafter, an embodiment of an imaging mass spectrometer using the mass spectrometry data analysis apparatus according to the present invention will be described with reference to the accompanying drawings. FIG. 1 is a schematic configuration diagram of an imaging mass spectrometer according to the present embodiment.

  This imaging mass spectrometer performs microscopic observation of a two-dimensional measurement region 8a on a sample 8 and executes an imaging mass analysis in the measurement region 8a, and an imaging mass analysis main unit 1 The data analysis processing unit 2 that performs analysis processing on the mass spectrometry imaging data collected in Step 1, the data storage unit 3 that stores mass analysis imaging data of the entire measurement region 8a, and the optical imaged by the imaging mass spectrometry main unit 1 A microscopic image processing unit 4 that processes the image signal to form a microscopic image, a control unit 5 that controls these units, and an operation unit 6 and a display unit 7 connected to the control unit 5 are provided.

Although not shown, the imaging mass spectrometry main body 1 includes, for example, a MALDI ion source, a three-dimensional quadrupole ion trap, a time-of-flight mass spectrometer, and the like, and measurement points (strictly speaking) set in the measurement region 8a. In other words, mass analysis over a predetermined mass-to-charge ratio range is performed on a microscopic area having a predetermined size. The imaging mass spectrometry main body 1 includes a drive unit that moves the stage on which the sample 8 is placed with high accuracy in two orthogonal x and y directions, and moves the sample 8 in the x-axis or y-axis direction. By executing mass spectrometry every time the sample is moved by a predetermined step length, it is possible to collect mass spectrum data for all measurement points in the measurement region 8a. In the imaging mass spectrometry main body 1, MS ⁿ analysis in which n is 2 or more can be performed by performing a selection operation of specific ions and a cleavage operation on the selected ions in the ion trap.

  In order to execute characteristic data analysis processing to be described later, the data analysis processing unit 2 includes a texture analysis unit 21, a feature amount spectrum creation unit 22, an analysis target peak determination unit 23, a data extraction unit 24, and a multivariate analysis processing unit. 25 etc. are included as functional blocks. Note that at least some of the functions such as the data analysis processing unit 2, the data storage unit 3, the microscopic image processing unit 4, and the control unit 5 execute dedicated processing / control software installed in a personal computer or workstation. Can be achieved.

The imaging mass spectrometer of the present embodiment analyzes and processes a huge amount of mass spectrometry imaging data in which mass spectrum data (or MS ⁿ spectrum data) for each measurement point obtained by the imaging mass spectrometry main body 1 is integrated. The data processing in the data analysis processing unit 2 displayed on the screen of the display unit 7 has its characteristics.
FIG. 2 is a schematic explanatory diagram of the data analysis processing operation in the imaging mass spectrometer of the present embodiment, and FIG. 3 is a flowchart showing a characteristic data analysis processing procedure in the imaging mass spectrometer of the present embodiment. 4 to 7 show results of actual measurement examples to which this data analysis process is applied. With reference to FIGS. 2 to 7, an example of data analysis processing that is a feature of the present invention will be described.

  At the time of performing analysis on the sample 8, the imaging mass spectrometry main body 1 moves in the x-axis direction, y in the measurement region 8a set two-dimensionally on the sample 8, as shown in FIGS. 2 (a) and 2 (b). Mass spectral data over a predetermined mass-to-charge ratio range is acquired for each minute region 8b finely divided in the axial direction. Note that the minute regions 8b, that is, the measurement points are not densely provided in the entire measurement region 8a as shown in FIG. 2, and may exist sparsely. That is, the pitch of the measurement points and the size of the measurement points can be arbitrarily determined.

  In general, in a mass spectrometry imaging image showing the spatial intensity distribution of ions having a certain mass-to-charge ratio, the display color on the color two-dimensional display image is determined for each minute region 8b. Accordingly, since the minute region 8b is the smallest unit in image processing such as coloring, the pixel in the image processing and the minute region are synonymous. In the following description, the minute region (measurement point) is defined as a pixel as necessary. Call it.

The data analysis process shown in the flowchart of FIG. 3 is executed in a state where the mass spectrometry imaging data for the measurement region 8 a on the sample 8 is stored in the data storage unit 3.
For example, when an analyst designates data to be analyzed from the operation unit 6 (for example, a file storing mass spectrometry imaging data for a target measurement region) and instructs the start of analysis, the data analysis processing unit 2 in the texture analysis unit 2 21 reads all the mass spectrum data in the designated measurement region from the data storage unit 3 (step S1). As shown in FIG. 2B, the mass spectrum at one certain measurement point (micro area or pixel) is composed of signal intensity values for each mass-to-charge ratio value. In the following description, a mass-to-charge ratio value at which a signal intensity value is obtained in a mass spectrum is referred to as a data point, and a serial number assigned to each data point when data points are arranged in ascending order of mass-to-charge ratio is referred to as a point number. I will do it.

The texture analysis unit 21 performs statistical texture analysis on an image composed of signal intensity value data at all measurement points included in the measurement region 8a for each data point.
Specifically, first, a point number smz corresponding to the lower limit value of the designated mass-to-charge ratio range is set as an initial value to a variable i indicating the point number of the mass spectrum data (step S2). Then, as shown in FIG. 2 (c), the signal strength value data whose point number is the variable i is collected from all the measurement points, and is reconstructed to correspond to the point number, that is, to the point number. Create an image for the corresponding mass to charge ratio value. A predetermined statistical texture analysis is performed on the image (step S3).

  In the statistical texture analysis at this time, paying attention to the density of the spatial intensity distribution of each mass-to-charge ratio value, feature quantities such as uniformity, directionality, and contrast change are calculated. For statistical texture analysis, several analysis methods such as a density histogram method (GLHM = Gray Level Histogram Method), a spatial density level dependent method, and a density level difference method are known. According to the examination, the spatial density level dependent method or the density level difference method is appropriate. In addition, various feature quantities can be obtained as a result of texture analysis. Here, in the case of the density level difference method, the correlation value is used as the feature quantity when the spatial density level dependence method is used. A contrast value may be used as the feature amount.

Since statistical texture analysis of an image is well known (see, for example, Non-Patent Documents 3 and 4), a detailed description thereof will be omitted and will be briefly described.
The spatial density level dependent method is a co-occurrence matrix whose element is the probability P (i, j) that the density of another pixel separated by a distance d in the angle θ direction from one pixel whose density is i in the image is j. This is a method for calculating the feature amount from the matrix. The correlation value a, which is one of the feature quantities, is a statistical measure of correlation weighted by the simultaneous occurrence probability of pixel pairs, and is obtained by the following equation.
a = Σ {(i−μ _i ) (j−μ _j ) P (i, j)} / σ _i σ _j
Here, Σ is the sum of i and j, and μ _i and σ _i are the mean and standard deviation of the probability P (i) that the concentration is i, μ _j , and σ _j are the concentrations of j, respectively. The average and standard deviation of the probability P (j). However, P (i) = Σ _j P (i, j), P (j) = Σ _i P (i, j), Σ _j is a sum related to j, and Σ _i is a sum related to i. When the image homogeneity is high (that is, the density change is small), the correlation value is low.

On the other hand, the density level difference method has a probability P (k) that the difference between the density of one pixel and the density of another pixel that is separated from the pixel by a distance d in the angle θ direction is k (where k = 0, 1, 2,..., N−1) is a method for calculating a feature value from the matrix. Contrast c, which is one of the feature quantities, is a scale indicating a local change in the matrix, and is obtained by the following expression.
c = Σk ² P (k)
Here, Σ is the sum of k from 0 to n-1.

  If a statistical texture analysis is performed on a certain variable i and a feature amount is obtained, it is determined whether or not the variable i is a point number emz corresponding to the upper limit value of the designated mass-to-charge ratio range. However, if i = emz, the texture analysis over the designated mass-to-charge ratio range is completed, and the process proceeds to step S6. On the other hand, if i = emz is not satisfied, the variable i is incremented (step S5) and the process returns to step S3. By repeating steps S3 to S5, the texture analysis can be performed on all the data points within the designated mass-to-charge ratio range, and the feature amount can be obtained.

  When the texture analysis ends, the feature amount spectrum creation unit 22 creates a feature amount spectrum in which the horizontal axis represents the mass-to-charge ratio m / z and the vertical axis represents the feature amount value, using the feature amount of each data point ( Step S6). As a result, a spectrum that is different from the above-described average spectrum and base peak spectrum and more accurately reflects the spatial distribution of the mass spectrometry imaging data can be obtained.

  Subsequently, the analysis target peak determining unit 23 performs peak detection on the feature amount spectrum obtained for only one target measurement region 8a, and extracts a peak (step S7). Then, the intensities (features) of the extracted many peaks are examined in order, a predetermined number of peaks are selected as the analysis target peaks in descending order of the intensities, and the mass-to-charge ratio values of the selected peaks are obtained (step S8). The number of peaks to be selected here, that is, the predetermined number is set as a predetermined value, is automatically determined in the course of processing, or an analyzer inputs and sets an arbitrary value from the operation unit 6. It may be determined.

  In step S8, the mass-to-charge ratio value from which data (signal intensity data) is to be extracted from the mass spectrum data obtained at each measurement point is determined. Therefore, the data extraction unit 24 extracts signal intensity data corresponding to the mass-to-charge ratio value determined in step S8 from the mass spectrum data obtained at each measurement point. Therefore, for example, when the number of peaks to be analyzed is 200, 200 different mass-to-charge ratio values are obtained, and 200 signal intensity data are obtained from each measurement point.

The data selected in this way is given to the multivariate analysis processing unit 25, and the multivariate analysis processing unit 25 performs a predetermined multivariate analysis process on the collected data so that it can be widely used in the measurement region 8a. A distributed substance or a substance localized in a specific part is discriminated (step S10). Finally, the multivariate analysis result is displayed on the screen of the display unit 7 to provide information to the analyst (step S11).
Hereinafter, an actual measurement example in which the data analysis processing described based on FIGS. 2 and 3 is applied to mass spectrometry imaging data obtained by actually performing analysis will be described.

[Measurement Example 1]
As an actual measurement example 1, an example in which data obtained by imaging mass spectrometry in positive ion mode using a slice of ginger sliced as a sample will be described.
FIG. 4 is a diagram showing the result of peak extraction at this time and a mass spectrometry imaging image at the mass-to-charge ratio value of each peak. Here, the average spectrum, base peak spectrum (these two are the conventional methods), and the feature amount spectrum created based on the correlation value calculated by the spatial concentration level dependent method (denoted as “SGLDM-corr spectrum” in the figure) ) Based on each, 200 peaks were selected as analysis target peaks in descending order of intensity.

  The numerical values in Fig. 4 are 200 analysis target peaks, each of which is selected from five mass-to-charge ratios of m / z 109.92, m / z 191.09, m / z 333.15, m / z 357.18, m / z 381.09. The number of the intensity ranking is indicated by “×”, and “×” indicates that a peak having the mass-to-charge ratio could not be selected (at least not in the top 200 intensities). The five types of mass-to-charge ratios correspond to substances that are known in advance to be distributed relatively specifically. Therefore, it can be said that the one in which these mass-to-charge ratios are selected is a spectrum reflecting a specific spatial distribution.

  FIG. 5 shows spectra used for selection of analysis target peaks. (A) is an average spectrum, (b) is a base peak spectrum, and (c) is a feature amount (SGLDM-Corr) spectrum. In each of these spectra, among the five kinds of mass-to-charge ratio values shown in FIG. 4, the selected peak is marked with a circle, and the peak that could not be selected is marked with a mark.

  As shown in FIG. 4, according to the result of the actual measurement example 1, when the SGLDM-Corr spectrum was used, it was not possible to select m / It can be seen that the peak at z 357.18 has been selected with the highest intensity ranking. Further, it can be confirmed that the peak at m / z 333.15 can be selected with higher intensity when the SGLDM-Corr spectrum is used than when the average spectrum or the base peak spectrum is used. When the mass spectrometry imaging image shown in FIG. 4 is seen, the mass-to-charge ratio selected in the SGLDM-corr spectrum, such as m / z 333.15 and m / z 357.18, is a space where pixels with relatively high intensity are localized. It can be seen that the distribution is shown. From this, it can be concluded that according to the data analysis method used in the apparatus according to the present invention, a spectrum reflecting the spatial distribution of the substance can be created as compared with the conventional method.

[Measurement Example 2]
As an actual measurement example 2, an example will be described in which data obtained by performing imaging mass spectrometry in positive ion mode using a slice of the skin of the sole of the mouse as a sample is analyzed.
FIG. 6 is a diagram showing the result of peak extraction at this time and a mass spectrometry imaging image at the mass-to-charge ratio value of each peak. As in the measurement example 1, based on the average spectrum, base peak spectrum, and SGLDM-corr spectrum, the top 200 peaks with the highest intensity were selected as analysis target peaks.

  FIG. 7 shows spectra used for selection of analysis target peaks. (A) is an average spectrum, (b) is a base peak spectrum, and (c) is an SGLDM-Corr spectrum. In these spectra, among the eight types of mass-to-charge ratio values shown in FIG. 6, the selected peak is marked with a circle, and the peak that could not be selected is marked with a mark.

  As shown in FIG. 6, according to the result of the actual measurement example 2, when the SGLDM-Corr spectrum was used, it was not possible to select m / It can be seen that peaks at z 630.4 and m / z 866.6 have been selected. When the mass spectrometry imaging image shown in FIG. 6 is seen, the mass-to-charge ratio selected in the SGLDM-corr spectrum is a space where pixels with relatively high intensity are localized, such as m / z 630.4 and m / z 866.6. It can be seen that the distribution is shown. From this, it can be concluded that according to the data analysis method used in the apparatus according to the present invention, a spectrum reflecting the spatial distribution of the substance can be created as compared with the conventional method.

It should be noted that the above embodiment is merely an example of the present invention, and it is obvious that modifications, corrections and additions may be made as appropriate within the scope of the present invention, and included in the claims of the present application.
For example, in the above embodiment, a characteristic data analysis process is applied to mass spectrometry imaging data obtained by collecting mass spectrum data obtained by ordinary mass spectrometry. However, a specific precursor ion is cleaved by collision-induced dissociation or the like. A characteristic data analysis process may be applied to mass spectrometry imaging data obtained by collecting MS ⁿ spectrum data obtained by mass analysis of the generated product ions.

DESCRIPTION OF SYMBOLS 1 ... Imaging mass spectrometry main-body part 2 ... Data processing part 21 ... Texture analysis part 22 ... Feature-value spectrum creation part 23 ... Analysis object peak determination part 24 ... Data extraction part 25 ... Multivariate analysis processing part 3 ... Data storage part 4 ... Microscopic image processing unit 5 ... control unit 6 ... operation unit 7 ... display unit

Claims (6)

A mass spectrometry data analyzer for analyzing mass spectrum data collected by executing mass spectrometry on each of a plurality of measurement points set in a two-dimensional measurement region on a sample,
a) Based on the mass spectrum data for each measurement point in the measurement region, texture analysis is performed for each image showing a two-dimensional distribution of signal intensity values created for each mass-charge value within a predetermined mass-to-charge ratio range. A texture analysis execution unit for obtaining a feature amount;
b) a feature amount spectrum creation unit that creates a mass spectrum based on the feature amount obtained for each mass-to-charge ratio value by the texture analysis execution unit;
c) an analysis target peak determination unit that performs peak detection on the feature amount spectrum created by the feature amount spectrum creation unit, and extracts a predetermined number of peaks as an analysis target peak in order of intensity among the obtained plurality of peaks; ,
A mass spectrometry data analysis apparatus comprising: selecting data from mass spectrum data at each measurement point in the measurement region with reference to the analysis target peak determined by the analysis target peak determination unit . The mass spectrometry data analysis device according to claim 1,
The mass analysis data analysis apparatus, wherein the texture analysis execution unit executes statistical texture analysis. The mass spectrometry data analysis device according to claim 2,
The mass analysis data analysis apparatus, wherein the texture analysis execution unit executes texture analysis by a spatial concentration level dependent method. The mass spectrometry data analysis device according to claim 3,
The texture analysis execution unit uses a correlation value as a feature amount. The mass spectrometry data analysis device according to claim 2,
The mass analysis data analysis apparatus, wherein the texture analysis execution unit executes texture analysis by a density level difference method. The mass spectrometry data analysis apparatus according to claim 5,
The texture analysis execution unit uses a contrast value as a feature amount.